Ideal fits in the span of five early waveforms left 50.8 percent of fixed SEEG forecast error at 20–500 ms in eight patient sessions
Five early stimulation responses defined a fixed waveform space, and five later responses tested it in eight patient sessions from OpenNeuro ds004696 v1.0.1. Ideal fits made after observing the later voltage left 50.8% of fixed forecast error (descriptive 95% patient bootstrap interval 47.1–54.7%); a fit in the early mean direction alone left 84.5%. The test used 886 stimulation-to-readout pairs at 86 sites. Four added directions captured 25.8% of later voltage energy after stimulation and 31.7% in a pre-pulse control (paired difference −5.8 percentage points, descriptive interval −10.1 to −1.5). These results bound error for this fixed early space on observed data. The fitted coefficients use the later voltage, so the test does not measure live forecast accuracy. The control limits a pulse-specific neural interpretation of the captured voltage.
SEEG · electrical stimulation · waveform projection · brain-computer interfaces · OpenNeuro
References
3- usesIdeal signed scale changes left 84.5 percent of 20–500 ms SEEG error for a fixed five-pulse waveform in eight patient sessions sci:sha256:c1013f1f75ee1d942a3ee7287ce6ea0c2a5f9632b77790eb3acd7cd28e98b3b6
We used its 886 exact ten-pulse Gram matrices, pulse selection, early energy weights, reference forecast, source records, and extraction code. We reproduced its 84.5 percent one-direction result and added a bound for the full five-waveform space.
- usesdoi:10.18112/openneuro.ds004696.v1.0.1
The study uses OpenNeuro ds004696 version 1.0.1 data and source metadata.
- doi:10.1523/JNEUROSCI.2201-22.2023
Cited by
2 publicationsPublic catalog entries on Sciscape.
- usesWith four fit pulses, mean SEEG shape change was 1.04 percentage points above a before-pulse change at 20–500 ms in eight people
Its waveform-space bound in another SEEG dataset informed the projection method and its observed-target limits. We add a held-out early pulse and a paired order/time change test in STReEF. We use no data or code from that package.
- usesFixed pre-pulse power weights did not show lower normalized SEEG mean forecast error at 20 to 500 ms with five calibration pulses in eight sessions
This study uses its exact ten-pulse voltage product matrices, fixed contact and site selection, source metadata, and source extraction code. The new test compares early pre-pulse power weights with the equal mean on later mean waveforms. It also derives and checks an exact weight assignment control.